Wireless mesh network access mode selection method and system based on timegan and multi-armed bandit

Abstract

The application discloses a wireless Mesh network access mode selection method and system based on TimeGAN and a multi-arm tiger machine, and the method comprises the following steps: forming a wireless Mesh network, each Mesh node broadcasts a data packet to other Mesh nodes in a high-frequency H frequency band to form a wireless network, and terminal access is performed in a low-frequency L frequency band; when a node joins or leaves the wireless Mesh network, a routing protocol adjusts the wireless Mesh network; a Mesh node is set as a controller, each Mesh node periodically scans a wireless environment, calculates available channel interference at the node and reports the same to the controller, the controller generates an adversarial network to obtain throughput data of various access strategies in the current wireless interference environment, and then uses the multi-arm tiger machine to determine optimal access strategies of each node and issues the optimal access strategies for execution. By monitoring a communication environment, the advantages and disadvantages of the wireless environment are determined according to interference generated by AP points in different channels in the environment, the optimal access mode is dynamically selected for the Mesh backbone network, and therefore the communication quality and wireless resource utilization rate are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of mobile communication, and particularly relates to a wireless Mesh network access mode selection method and system based on TimeGAN and a multi-armed tiger machine. BACKGROUND

[0002] With the rapid development of mobile Internet and wireless communication technology, people's lifestyle has been greatly changed. Various network services have improved people's life experience. However, users begin to pursue better communication services and experience, and need multimedia mobile communication to provide more diverse service categories. Therefore, many high-speed broadband communication technologies have emerged. As a new type of Internet access technology, wireless Mesh network provides convenient and economical Internet access services for users.

[0003] Wireless Mesh network is a multi-hop network, which is different from traditional wireless network, and combines the advantages of wireless local area network (WLAN) and Ad-hoc network, and has the advantages of high speed, easy networking, strong robustness and self-organization. In practice, wireless Mesh network has been widely applied, for example, wireless campus network, wireless city, tourist and leisure place, video monitoring, etc. can adopt wireless Mesh network for networking.

[0004] The traditional wireless Mesh network will cause the communication quality to decrease due to the interference of other AP points in the environment, and in addition, the unreasonable channel bandwidth, transmission power and other access modes of the Mesh node will also cause the problems of low communication quality and low wireless resource utilization.

[0005] Therefore, it is necessary to design a new wireless Mesh network access mode to solve the above problems. SUMMARY

[0006] The main purpose of the application is to monitor the communication environment of the Mesh backbone network, determine the advantages and disadvantages of the wireless environment according to the interference generated by the AP points in different channels in the environment, help the Mesh backbone network to dynamically select the optimal access mode, and thus improve the communication quality and wireless resource utilization.

[0007] To achieve the above purpose, the application provides a wireless Mesh network access mode selection method based on TimeGAN and a multi-armed tiger machine, which comprises the following steps:

[0008] Step one, a wireless Mesh network is established, and each Mesh node broadcasts data packets to other Mesh nodes in a high-frequency H frequency band for wireless networking, and performs terminal access in a low-frequency L frequency band;

[0009] Step two, when a node joins or leaves the wireless Mesh network, the routing protocol will adjust the wireless Mesh network;

[0010] Step three, set a Mesh node as a controller, each Mesh node periodically scans the wireless environment, calculates the interference of available channels at each node and reports to the controller, the controller generates the current wireless interference environment through time series to obtain the throughput data of various access strategies, and then uses a multi-arm tiger machine to make the optimal access strategy of each node and issues it to each node for execution.

[0011] Further improvement of the application is that step one further comprises the following steps:

[0012] Step S1.1, each Mesh node periodically broadcasts a neighbor discovery data packet, the content of the data packet includes link quality, TTL value, last node MAC address, sequence number, source node MAC address and data packet type, the broadcast data packet is used to synchronize the transmission link quality of each node, the source node informs other nodes of the existence of itself and the next hop node that can be reached by itself;

[0013] Step S1.2, the Mesh node updates the routing table according to the received broadcast data packet of other node neighbor discovery, and the main update content is: the source node chain table that can be reached by the local Mesh node and the link quality between the local Mesh node and the one-hop neighbor;

[0014] Step S1.3, set the name of the Mesh backbone network, and set the transmission power, channel and channel bandwidth for each node, and build the Mesh backbone network in the H frequency band.

[0015] Further improvement of the application is that step two further comprises the following steps:

[0016] Step S2.1, when a node suddenly joins the Mesh backbone network, it will send a broadcast data packet of neighbor discovery to other nodes, and receive the broadcast data packet of other Mesh nodes, and after updating the routing table of each Mesh node, successfully join the Mesh network;

[0017] Step S2.2, when the node fails or leaves the network, the transmission link quality between the node and other nodes becomes 0, all nodes will automatically avoid the node when routing, and the whole wireless Mesh network continues to work normally.

[0018] Further improvement of the application is that step three further comprises the following steps:

[0019] Step 3.1, there are multiple optional channels in the H frequency band of the Mesh node wireless networking, each node periodically scans the wireless environment, and the current wireless channel environment is represented based on the signal reception interference;

[0020] Step 3.2, set a Mesh node as a controller, collect the wireless environment reported by other Mesh nodes, and generate the throughput obtained by various access strategies in the current wireless environment through the model trained by TimeGAN; take the channel environment Ej of the node, the selected channel Chj, the used channel bandwidth BWj, the node transmission power Powj, and the throughput TPj at the node as samples, collect sample data under different wireless environments in the Mesh backbone network, and train the data through TimeGAN to obtain a model, input the wireless environment at each Mesh node into the model, and output the throughput of the corresponding Mesh node under different access strategies in the current environment, and the format is {Chj, BWj, Powj, TPj}.

[0021] Step 3.3, using {Chj, BWj, Powj} as action and {TPj} as reward, scoring each access strategy of each node generated by TimeGAN using a multi-armed bandit, and deciding the optimal access strategy under the current unlimited environment at each node; and

[0022] Step 3.4, after the controller completes the decision, the results are respectively issued to the corresponding Mesh node, and after the Mesh node receives the access decision issued by the controller, the corresponding access strategy is deployed according to the needs on the premise of not affecting the communication quality.

[0023] Further improvement of the application is that the step 3.1 further comprises the following steps:

[0024] Step 3.1.1, the Mesh node scans the wireless environment and counts the received AP point information: the channel where the AP point is located and the available bandwidth of the channel, the received signal strength RSSI of the AP point at the node, and calculates the interference of the AP point on the corresponding wireless channel at the node: Pt=10^(RSSI / 10);

[0025] Step 3.1.2, according to the interference of the AP point on the corresponding wireless channel at the node, calculate the interference on different channels at the Mesh node;

[0026] Step 3.1.3. The total channel power Pi of each channel at the node can be calculated by the AP point information interfering with the node, and the channel weighted power Pwi reflecting the interference between channels needs to be calculated, which can reflect the interference degree of each channel in the wireless environment at the node, and the smaller the channel weighted power Pwi is, the smaller the interference of the channel is, and the Pwi through the channel can represent the wireless environment of the node.

[0027] Further improvement of the application is that the interference of the AP point to the environment channel at the Mesh node is P i i =∑k t ×P t (k t =0,0.25,0.5,1), wherein kt represents the number of channels interfering with the AP point, and kt represents the interference between adjacent channels determined by the channel bandwidth.

[0028] Further improvement of the application is that the weighted channel power Pwi:

[0029]

[0030] As a further improvement of the application, the wireless Mesh network can be based on the IEEE 802.11 protocol, and the networking can be carried out in the 5GHz frequency band, and the terminal access can be carried out in the 2.4GHz frequency band.

[0031] To achieve the above application purposes, the application further provides a wireless Mesh network access mode selection system based on TimeGAN and multi-arm tiger machine, which can implement the method as described in any one of the preceding embodiments.

[0032] The beneficial effects of the application are as follows:

[0033] (1) The application can realize real-time sensing of the interference generated by different AP points in the environment by scanning the channel environment, and can adaptively select the optimal access strategy at each node to ensure the throughput and delay of the Mesh network.

[0034] (2) The application can realize adaptive switching of the access mode according to the interference of the wireless environment by setting the controller to control the entire Mesh network.

[0035] (3) The Mesh backbone network designed by the application runs in the data link layer, and the routing information is transmitted in the form of Ethernet frames, which enhances the stability of the network and reduces the delay. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is an effect diagram of the wireless Mesh network access mode selection method based on TimeGAN and multi-arm tiger machine of the application; ​

[0037] Figure 2 is the flow chart of the Mesh network building process of the present application;

[0038] Figure 3 is the flow chart of the Mesh node sensing the interference condition of the wireless environment and reporting to the controller of the present application;

[0039] Figure 4 is the flow chart of the controller making decisions according to the reports of the Mesh nodes to the controller of the present application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be described in detail below in combination with the drawings and specific embodiments.

[0041] It should be emphasized that in the description of the present application, various formulas and constraints are distinguished by consistent labels before and after, but different labels may also be used to mark the same formulas and / or constraints. The purpose of this setting is to make the features of the present application clearer.

[0042] The present application provides a wireless Mesh network access mode selection method based on TimeGAN and multi-armed bandit. The method mainly includes the following steps:

[0043] Step one, build a wireless Mesh network, each Mesh node broadcasts data packets to other Mesh nodes in the H frequency band with higher frequency for wireless networking, and in the L frequency band with lower frequency for terminal access;

[0044] Step two, when a node joins or leaves the wireless Mesh network, the routing protocol will adjust the wireless Mesh network;

[0045] Step three, set a Mesh node as a controller, each Mesh node periodically scans the wireless environment, calculates the interference condition of the available channel at each node and reports it to the controller, the controller obtains the throughput data of various access strategies in the current wireless interference environment through the time series generation adversarial network, and then uses the multi-armed bandit to make the optimal access strategy for each node and issues it to each node for execution.

[0046] The following will describe each step in detail in combination with the embodiments.

[0047] For example, Figure 1As shown, the application provides a wireless Mesh network access mode selection method based on TimeGAN and multi-armed bandit. A network controller is added in the traditional wireless Mesh network, and the whole Mesh network is controlled through the method of software-defined network. Each Mesh node scans the wireless environment, calculates the interference of the available channels at the Mesh node and reports it to the controller, the controller obtains the throughput data of various access strategies under the current wireless interference environment through the time series generation adversarial network, and then uses the multi-armed bandit to decide the optimal access strategy of each node and issues it to each node for execution, thereby ensuring the communication quality of the Mesh backbone network.

[0048] The Mesh network building process described in the application is as shown in Figure 2 .

[0049] Step S1, the wireless Mesh network is established, each Mesh node broadcasts data packets to other Mesh nodes in the H frequency band with higher frequency to perform wireless networking, and performs terminal access in the L frequency band with lower frequency. For example, networking can be performed in the 5GHz frequency band based on the IEEE802.11 protocol, and terminal access can be performed in the 2.4GHz frequency band, which specifically includes the following steps:

[0050] Step S1.1, each Mesh node periodically broadcasts a data packet for discovering neighbors, and the content of the packet includes link quality, TTL value, last node MAC address, sequence number, source node MAC address, packet type and the like. The broadcast data packet has two functions: synchronizing the transmission link quality of each node, and notifying other nodes of the existence of the source node and the next hop node reachable by the source node.

[0051] Step S1.2, the Mesh node updates the routing table according to the broadcast data packet for discovering neighbors received by other nodes, and the update content mainly includes: a source node chain table reachable by the local Mesh node and the link quality between the local Mesh node and the one-hop neighbor.

[0052] Step S1.3, setting the name of the Mesh backbone network, and setting the transmission power, channel and channel bandwidth for each node, and building the Mesh backbone network in the H frequency band.

[0053] Step S2, when a node suddenly joins or leaves the Mesh backbone network, the routing protocol quickly adjusts the Mesh backbone network. Specifically:

[0054] Step S2.1, when a node suddenly joins the Mesh backbone network, it sends a broadcast data packet for discovering neighbors to other nodes, and receives the broadcast data packets of other Mesh nodes, and after updating the routing table of each Mesh node, successfully joins the Mesh network.

[0055] Step S2.2: When a node fails or leaves the network, the transmission link quality between the node and other nodes becomes 0. All nodes will automatically avoid the node when routing, and the entire wireless mesh network will continue to work normally.

[0056] The process by which the Mesh node senses wireless environmental interference and reports it to the controller as described in this invention is as follows: Figure 3 As shown.

[0057] Step A1: The H-band of the Mesh node wireless network has multiple selectable channels. Each node periodically scans the wireless environment and represents the current wireless channel environment based on the interference situation at the signal receiver. Specifically:

[0058] Step A1.1: The Mesh node scans the wireless environment and collects information about the access points (APs): the channel where the AP is located, the available bandwidth of that channel, and the received signal strength (RSSI) of the AP at the node. Using this information, the interference caused by the AP to the corresponding wireless channel at the node is calculated: Pt = 10^(RSSI / 10).

[0059] Step A1.2: An access point (AP) can interfere with multiple channels on which a node is located. This is related to the AP's channel and channel bandwidth. Based on the interference caused by the AP to the corresponding wireless channels at the node, calculate the interference experienced by different channels at the Mesh node.

[0060] The interference of the AP point to the environmental channel at the Mesh node is P. i :P i =∑k t ×P t (k t =0,0.25,0.5,1), where kt represents the number of channels that cause interference at this AP point, which is determined by the channel bandwidth to indicate the interference situation between adjacent channels.

[0061] Step A1.3: Using the AP information that causes interference to the node, the total channel power Pi of each channel at that node can be calculated. However, Pi cannot fully reflect the interference situation of the channel because interference can also occur between adjacent channels. It is necessary to calculate the weighted channel power Pwi, which reflects the interference between channels:

[0062]

[0063] Channel weighted power (Pwi) can reflect the degree of interference of each channel in the wireless environment at a node. The smaller the channel weighted power, the less interference the channel receives. The Pwi of these channels can represent the wireless environment of the node.

[0064] Step A2: Set up a Mesh node as a controller, collect the wireless environment reported by other Mesh nodes, and generate the throughput obtained by various access strategies under the current wireless environment through the model trained by TimeGAN.

[0065] The node's channel environment Ej, selected channel Chj, channel bandwidth BWj, node transmit power Powj, and throughput TPj are used as samples. Sample data under different wireless environments are collected in the Mesh backbone, and a model is obtained by training these data using TimeGAN. Inputting the wireless environment of each Mesh node into this model, it can output the throughput of the corresponding Mesh node under different access strategies in the current environment, in the format {Chj, BWj, Powj, TPj}.

[0066] Step A3: Using {Chj, BWj, Powj} as actions and {TPj} as rewards, use a multi-armed slot machine to score each access strategy of each node generated by TimeGAN, and determine the optimal access strategy for each node in the current infinite environment.

[0067] Step A4: After making a decision, the controller sends the result to the corresponding Mesh nodes. Upon receiving the access decision from the controller, the Mesh nodes deploy the corresponding access strategy as needed, without affecting communication quality.

[0068] Based on the above inventive concept, the present invention also discloses a wireless mesh network access mode selection system based on TimeGAN and multi-armed slot machine, including at least one computing device. The computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded onto the processor, it can implement the above-mentioned wireless mesh network access mode selection method based on TimeGAN and multi-armed slot machine.

[0069] This invention scans the channel environment to detect interference from different access points in real time, adaptively selecting the optimal access strategy for each node to ensure the throughput and latency of the mesh network. Simultaneously, this invention manages the entire mesh network through a controller, enabling adaptive switching of access methods based on wireless interference conditions. Furthermore, the mesh backbone designed in this invention operates at the data link layer, transmitting routing information via Ethernet frames, enhancing network stability and reducing latency.

[0070] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for selecting access mode in a wireless mesh network based on TimeGAN and multi-armed slot machines, characterized in that: Includes the following steps: Step 1: Build a wireless mesh network. Each mesh node broadcasts data packets to other mesh nodes in the higher frequency H band to form a wireless network, and uses the lower frequency L band for terminal access. Step 2: When a node joins or leaves the wireless mesh network, the routing protocol will adjust the wireless mesh network. Step 3: Set up a Mesh node as a controller. Each Mesh node periodically scans the wireless environment, calculates the interference of available channels at each node and reports it to the controller. The controller obtains the throughput data of various access strategies under the current wireless interference environment through a time series generative adversarial network, and then uses a multi-armed slot machine to determine the optimal access strategy for each node and sends it to each node for execution. Includes the following steps: Step 3.1: The H band of the Mesh node wireless network has multiple selectable channels. Each node periodically scans the wireless environment and represents the current wireless channel environment based on the interference situation at the signal receiving end. Step 3.2: Set up a Mesh node as a controller to collect the wireless environment reported by other Mesh nodes, and generate the throughput obtained by various access strategies under the current wireless environment through the model trained by TimeGAN; using the node's channel environment Ej, selected channel Chj, channel bandwidth BWj, node transmit power Powj, and throughput TPj at the node as samples, collect sample data under different wireless environments in the Mesh backbone, and train these data through TimeGAN to obtain a model. Input the wireless environment at each Mesh node into the model, and it can output the throughput of the corresponding Mesh node under different access strategies under the current environment, in the format {Chj, BWj, Powj, TPj}. Step 3.3: Using {Chj, BWj, Powj} as actions and {TPj} as rewards, use a multi-armed slot machine to score each access strategy of each node generated by TimeGAN, and determine the optimal access strategy for each node in the current infinite environment. as well as Step 3.4: After the controller completes the decision, it sends the results to the corresponding Mesh nodes. After receiving the access decision from the controller, the Mesh nodes deploy the corresponding access strategies as needed without affecting the communication quality.

2. The method according to claim 1, characterized in that: Step one further includes the following steps: Step S1.1: Each Mesh node periodically broadcasts data packets to discover its neighbors. The data packets include link quality, TTL value, MAC address of the previous node, sequence number, MAC address of the source node, and data packet type. This broadcast data packet is used to synchronize the transmission link quality of each node, and for the source node to announce its own existence and the next hop node it can reach to other nodes. Step S1.2: The Mesh node updates its routing table based on the broadcast packets discovered by other neighboring nodes. The main updates are: the list of source nodes reachable by the local Mesh node and the link quality between the local Mesh node and its one-hop neighbor. Step S1.3: Set the name of the Mesh backbone network and set the transmit power, channel and channel bandwidth for each node to build the Mesh backbone network in the H band.

3. The method according to claim 1, characterized in that: Step two further includes the following steps: Step S2.1: When a node suddenly joins the Mesh backbone, it will send a neighbor discovery broadcast data packet to other nodes and receive broadcast data packets from other Mesh nodes. After each Mesh node updates its routing table, it will successfully join the Mesh network. Step S2.2: When a node fails or leaves the network, the transmission link quality between the node and other nodes becomes 0. All nodes will automatically avoid the node when routing, and the entire wireless mesh network will continue to work normally.

4. The method according to claim 1, characterized in that: Step 3.1 further includes the following steps: Step 3.1.1: The Mesh node scans the wireless environment and collects the received AP information: the channel where the AP is located and the available bandwidth of the channel, the received signal strength RSSI of the AP at the node, and calculates the interference caused by the AP to the corresponding wireless channel at the node: Pt=10^(RSSI / 10). Step 3.1.2: Calculate the interference experienced by different channels at the Mesh node based on the interference generated by the AP point on the corresponding wireless channel at the node; Step 3.1.3: The total channel power Pi of each channel at the node can be calculated using the AP information that causes interference to the node. It is necessary to calculate the channel weighted power Pwi, which reflects the interference between channels. The channel weighted power Pwi can reflect the degree of interference of each channel in the wireless environment at the node. The smaller the channel weighted power, the less interference the channel is subjected to. The Pwi of these channels can represent the wireless environment of the node.

5. The method according to claim 4, characterized in that: The interference of the AP point on the environmental channel at the Mesh node is as follows: ： （ =0,0.25,0.5,1）， Where kt represents the number of channels that cause interference at this AP point, and is determined by the channel bandwidth to indicate the interference situation between adjacent channels.

6. The method according to claim 5, characterized in that: The weighted channel power is Pwi: 。 7. The method according to claim 1, characterized in that: Wireless Mesh networks can be networked in the 5GHz band based on the IEEE 802.11 protocol, and terminal access can be achieved in the 2.4GHz band.

8. A wireless mesh network access mode selection system based on TimeGAN and multi-armed slot machine, characterized in that: The device includes at least one computing device, the computing device including a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being loaded onto the processor to perform the method as described in any one of claims 1 to 7.

Images